Highly diastereoselective and enantioselective preparation of homoallylic amines: Application for the synthesis of β-amino acids and γ-lactams

Article abstract of DOI:10.1002/chem.200401295

Reactions of N-silyl- and N-aluminoimines with B- allyldiisopinocampheylborane in the presence of methanol, followed by oxidative workup furnished homoallylic amines in good yields and high ee. A ¹¹B NMR spectroscopy study revealed that the reactions do not proceed, even at room temperature, unless a molar equivalent of water or methanol is added. The first reagent-controlled asymmetric crotylboration and alkoxyallylboration of aldimines furnishing β-methyl or βalkoxy homoallylic amines in very high diastereoselectivity and enantioselectivity are reported herein. Crotylboration and alkoxyallylboration of imines proceed only with the "allyl"-boron "ate" complexes, instead of the "allyl"-dialkylboron reagents used with aldehydes. The addition of methanol is necessary for these reactions as well. Application of this methodology for the conversion of representative nitriles to β-amino acids in two steps has been described. Additionally, a procedure for the preparation of chiral δ-amino alcohols and γ-lactams from nitriles is also reported.

Full text of DOI:10.1002/chem.200401295

FULL PAPER
Highly Diastereoselective and Enantioselective Preparation of Homoallylic
Amines: Application for the Synthesis of b-Amino Acids and g-Lactams
P. Veeraraghavan Ramachandran* and Thomas E. Burghardt[a]
Dedicated to Professor Ei-ichi Negishi on the occasion of his 70th birthday
Abstract: Reactions of N-silyl- and N-
aluminoimines with B-allyldiisopino-
campheylborane in the presence of
methanol, followed by oxidative
workup furnished homoallylic amines
in good yields and high ee. A 11B NMR
spectroscopy study revealed that the
reactions do not proceed, even at room
temperature, unless a molar equivalent
of water or methanol is added. The
first reagent-controlled asymmetric
crotylboration and alkoxyallylboration
of aldimines furnishing b-methyl or b-
alkoxy homoallylic amines in very high
diastereoselectivity and enantioselec-
tivity are reported herein. Crotylbora-
tion and alkoxyallylboration of imines
proceed only with the “allyl”-boron
“ate” complexes, instead of the “allyl”-
dialkylboron reagents used with alde-
hydes. The addition of methanol is nec-
essary for these reactions as well. Ap-
plication of this methodology for the
conversion of representative nitriles to
b-amino acids in two steps has been de-
scribed. Additionally, a procedure for
the preparation of chiral d-amino alco-
hols and g-lactams from nitriles is also
reported.
Keywords: allylic compounds
amines asymmetric synthesis
boranes · imines
·
·
·
Introduction
sulfiniyl,[4] N-sulfonyl,[5] N-trialkylsilyl,[6] oximes,[7] and vari-
ous N-metalloimines.[8] N-silylimines have been first pre-
pared by Rochow and co-workers by reacting aromatic alde-
hyde with lithium bis(trimethylsilyl)amide.[9] Reactions of
such imines with organometallic reagents for the prepara-
tion of homoallylic amines was first reported by Hart and
co-workers.[10] Enolizable N-silylimines are not stable, but
were characterized at very low temperatures (ꢀ1008C).[11]
Their preparation and in situ trapping at low temperatures
furnished only low yields of the synthesized b-lactam prod-
ucts.[12] This inefficiency of N-silylimines, however, could be
alleviated by the use of N-aluminoimines, which were first
prepared by Cainelli and co-workers by partial reduction of
nitriles with diisobutylaluminum hydride (DIBAL-H).[13]
Among the advantages of N-aluminoimines are the ready
availability or ease of preparation of the starting nitriles,
ease of the imine formation, and relatively high stability of
both aromatic and aliphatic imines, even at room tempera-
ture. Due to the above advantages of N-aluminoimines and
our ongoing interest in organoaluminum chemistry,[14] they
became a very interesting substrate for our investigation.
In spite of aldimines being very attractive starting materi-
als for syntheses, the inherent instability of unsubstituted al-
dimines and the lack of reactivity of the more stable N-sub-
stituted ones significantly limit their use.[15] Furthermore, at-
Preparation of optically pure homoallylic amines is an im-
portant task in organic synthesis. Such amines are excellent
building blocks for the synthesis of a plethora of nitrogen-
containing natural products.[1] Owing to the importance of
homoallylic amines, there are abundant literature reports
for their preparation using a wide variety of methods.[1d,2]
Diastereoselective addition of allyl organometallic com-
pounds to N-substituted imines is a widely recognized proce-
dure for the preparation of homoallylic amines.[3] While the
use of substrate-controlled asymmetric additions is
a
common procedure, the use of reagent control still remains
undeveloped.[3] Numerous N-substituted imines have been
developed and applied for the syntheses of a variety of mol-
ecules, such as amino acids, b-lactams, heterocycles, aziri-
dines, alkaloids, and amines.[3] These aldimines include N-
[a] Prof. P. V. Ramachandran, T. E. Burghardt
Herbert C. Brown Center for Borane Research
Department of Chemistry, Purdue University
560 Oval Dr., West Lafayette, IN 47907-2084 (USA)
Fax : (+1)765-494-0239
E-mail: chandran@purdue.edu
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2005, 11, 4387 – 4395
DOI: 10.1002/chem.200401295
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4387

tempted 1,2-additions to aldimines quite often fail due to
the tendency of enolizable compounds to undergo deproto-
nation instead of addition.[3e] Among the developments in
allylations of aldimines, organoboron compounds have
found synthetic utility due to the Lewis acidity of boron,
which coordinates to nitrogen and improves the electrophi-
licity of the aldimine. Allylborations and crotylborations of
a-optically pure N-substituted aldimines in high diastereose-
lectivity due to the sterically influenced chirality according
to the Cram rules, using B-“allyl”-9-BBN, was described by
Yamamoto and co-workers.[16] Reactions of arylimines,
oximes and oxime ethers with allyl and crotylboronates
were reported to proceed in moderate to good diastereose-
lectivity.[7] However, use of N-borylimine containing a chiral
auxiliary, followed by the addition of alkyllithium reagents
gave only poor enantioselectivity.[17]
Asymmetric allylboration, crotylboration, and alkoxyallyl-
boration of aldehydes with reagents I–IV derived from a-
pinene[18] (Figure 1) are well documented and widely used in
organic syntheses.[19] In spite of a few reports regarding
chiral allylboration of N-masked imines,[20] reactions of
these and other boron-based chiral reagents with imines to
form homoallylic amines remain significantly underdevel-
oped. Itsuno and co-workers had reported the preparation
of assorted homoallylic amines in moderate to good enantio-
selectivities from various N-protected imines, such as N-
oxime ethers, N-sulfenimines, and N-trimethylsilylimines, by
using chiral allylboranes derived from tartrate esters, diols,
a-amino alcohols, and a-pinene.[20] They concluded that N-
trimethylsilylimines are the most reactive species for such
allylations. During our work on the applications of a-
pinene-based “allyl”-borating reagents we always obtained
homoallylic alcohols in very high ee during the allylboration
of aldehydes.[19] However, the enantioselectivity reported for
the allylboration of N-trimethylsilylbenzaldimine (1a) with
I, complete within 3 h, was only a moderate 73%.[20a] Since
the rate of allylboration of aldehydes with I was established
as exceptionally fast at ꢀ788C, and fast even at ꢀ1008C,[21]
it appeared desirable to obtain comparable information
about the allylboration of imines. The low ee value and the
slow rate of the reaction prompted us to undertake a project
involving “allyl”-boration of N-substituted aldimines. The
preliminary results were reported earlier.[22]
Results and Discussion
Allylboration of N-silylimines: N-Trimethylsilylbenzaldimine
(1a) was mixed with (ꢀ)-B-allyldiisopinocampheylborane
(I) at ꢀ788C in THF and the reaction was monitored with
11B NMR spectroscopy. The spectrum of an aliquot revealed
only unchanged starting materials, even after several hours
at room temperature. This was a surprise since allylboration
of aldehydes with I was previously established as an ex-
tremely fast reaction[21] and the previous report[20a] claimed
that the reaction was complete within 3 h. However, the de-
sired amine 2a was obtained in good yield after aqueous
workup, which was very exothermic. We rationalized that
the reaction must have taken place during the workup when
the addition of water resulted in the liberation of the
“naked” aldimine intermediate, which rapidly reacted with
I. We were not able to identify this intermediate spectro-
scopically, presumably due to the extremely fast rate of the
allylboration.[21] Consequently, 1a was mixed with I at
ꢀ788C in THF and upon dropwise addition of 1 equivof
water dissolved in THF to the reaction mixture and workup
provided 2a in 90% yield and 92% ee, which was considera-
bly better than 70% yield and 73% ee previously reported
(Scheme 1).[20a] Lowering the reaction temperature to
ꢀ1008C increased the chiral induction to 94% ee.
Scheme 1. Allylboration of N-silylimines: a) I, THF; ꢀ100 or ꢀ788C;
b) 1a–e; c) H2O; 1 h, ꢀ100 or ꢀ788C; d) NaOH, H2O2; ꢀ788C!RT.
Thp = thiophene.
After we had standardized the reaction conditions, we ex-
amined the allylboration of other N-silylimines prepared
from aromatic aldehydes, namely 2-thienyltrimethylsilyli-
mine (1b), 4-methoxybenzaldimine (1c), 2-chlorobenzaldi-
mine (1d), and 2-furfuraldimine (1e) (Scheme 1). The
above N-silylaldimines upon reaction with I in the presence
of 1 equivof water, followed by workup furnished the ex-
pected homoallylic amines 2b–e in good yields and high
enantioselectivity (Table 1). The yields and enantioselectivi-
ties of 2b and c were considerably better than previously re-
ported.[20a] We thus established that N-silylimines were not
reactive toward I unless a molar equivalent of water was
added to the reaction mixture.[22] Later studies revealed that
methanol could replace water.[23]
Crotylboration and alkoxyallylboration of N-silyl- or N-
aluminoimines for the preparation of densely functionalized
homoallylic amines have never been reported. In this paper,
we describe the results of our investigations on the “allyl”-
boration of N-silyl- and N-aluminoimines.
Crotylboration of N-silylimines: Crotylboration of aldehydes
with II and III[18c] is a well-known and highly utilized proce-
dure for the syntheses of numerous complex natural prod-
ucts.[19,24] However, crotylboration of N-substituted imines
Figure 1. a-Pinene-based reagents for asymmetric allylboration.
4388
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4387 – 4395

Homoallylic Amines
FULL PAPER
Table 1. Allylboration of N-silylimines.
amine 3a in 78% yield, 92% ee
and 96% de. The reaction of VI
with 1a, after the addition of
methanol was also complete
within 3 h and alkaline oxida-
tive workup, followed by purifi-
cation provided the amine 4a in
76% yield, 90% ee, and 98%
de (Scheme 2).
Entry
Imine
R
T [8C]
Homoallylic amine
Yield [%][a]
ee [%][b]
1
1
2
3
4
5
1a
1a
1b
1c
1d
1e
Ph
Ph
2-Thp
4-MeO-(C6H4)
2-Cl-(C6H4)
2-furyl
ꢀ78
ꢀ100
ꢀ78
ꢀ78
ꢀ78
ꢀ78
2a
2a
2b
2c
2d
2e
90
87
72
74
69
86
92
94
81
92
82
86
[a] All yields are of pure isolated products. [b] Enantiomeric excess was determined with HPLC using Chiracel
OD-H column and hexanes/isopropanol as the mobile phase.
with these reagents has never been reported. Unlike I,
which can be prepared and stored for short periods of time,
the reagents II and III are made fresh and used immediate-
ly.[18c] The treatment of cis- or trans-butene with Schlosserꢁs
base (equimolar mixture of butyllithium and potassium tert-
butoxide) at ꢀ558C furnishes the crotyl anion.[25] Reaction
of this anion with B-methoxydiisopinocampheylborane fur-
nishes the corresponding boron “ate” complex (V or VI). In
the crotylboration of aldehydes, this “ate” complex is
broken by the addition of 1.3 equivBF 3·OEt2 to generate
trialkylborane reagents II or III (Figure 2).[18c,26] However,
we found that addition of BF3·OEt2 had detrimental effect
on crotylboration of N-silylaldimines and multiple unidenti-
fied products were obtained. Indeed, a blank experiment, in
which 1a was mixed with methanol and BF3·OEt2 showed
Scheme 2. Crotylboration of N-silylimines.
Alkoxyallylboration of N-silylimines: Alkoxyallylboration of
aldehydes with IV[18d] has found many applications in organ-
ic syntheses.[19,28] However, there are no literature reports
regarding alkoxyallylboration of imines for the preparation
of b-alkoxy homoallylic amines. Similar to the reagents II
and III, alkoxyallylboration reagent IV is prepared freshly
before reaction. Treatment of allyl ether with sec-butyllithi-
um furnishes the (Z)-allylic anion due to the coordination
between the oxygen and lithium.[18d] The allylic anion reacts
with B-methoxydiisopinocampheylborane to give the corre-
sponding boron “ate” complex VII, which upon the addition
of BF3·OEt2 gives the reagent IV (Figure 3).[18d] However,
again, as in the case of crotylboration of N-silylimines, the
reactions were not compatible with BF3·OEt2 and the “ate”
complex VII was used for alkoxyallylboration.
1
immediate degradation of 1a (based on a H NMR analysis
of an aliquot). Therefore, we resorted to crotylboration of
the aldimines with the “ate” complexes V and VI (Figure 2).
Figure 2. Preparation of reagents for asymmetric crotylboration. Ipc =
isopinocampheyl.
To the best of our knowledge, this is the first report of
crotylboration with “ate” complexes V or VI. We had re-
cently reported similar alkoxyallylboration of fluoral with
“ate” complex VII.[27] We do not know the exact mechanism
of the crotylboration with the “ate” complex. The 11B NMR
spectrum of the reaction mixture containing 1a and V
showed a peak at d 4, corresponding to the “ate” com-
Figure 3. Preparation of alkoxyallylborane reagents.
1
plex[18c] and in H NMR spectrum there was a peak at d 9.0,
corresponding to the unreacted N-silylimine. Upon addition
of methanol, we observed an exothermic reaction and
11B NMR spectrum of an aliquot showed the formation of a
broad peak at d 48 and disappearance of the peak at d 4.
Additionally, we observed the disappearance of the peak at
d 9.0 in the crude H NMR spectrum. We are currently in-
vestigating the mechanism in detail.
The reaction was complete within 3 h and alkaline oxida-
tive workup provided the expected b-methyl homoallylic
Our experience in alkoxyallylboration revealed that the
best results were obtained with 3-[(2-methoxyethoxy)me-
thoxy]prop-1-ene (allyl-OMEM; R = CH2OCH2CH2OCH3)
due to the ease of hydrolysis of the MEM group.[28a] Accord-
ingly, this reagent was evaluated for the preparation of the
b-alkoxy homoallylic amine. Thus, when 1a and VII were
mixed in THF at ꢀ788C, followed by the addition of 1 equiv
methanol, upon completion of the reaction (3 h, monitored
with 11B NMR spectroscopy), oxidative workup furnished
1
Chem. Eur. J. 2005, 11, 4387 – 4395
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4389

P. V. Ramachandran and T. E. Burghardt
at ꢀ788C (Scheme 4; Table 2, entry 1).[29] Thus, we con-
firmed that methanol is a critical additive in the allylbora-
tion of N-aluminoimines as well. This finding and the spec-
trometric data support the postulated mechanism (Figure 4).
Scheme 3. Alkoxyallylboration of N-silylimines.
the expected amine 5a in 65% yield, 95% ee, and 98% de
(Scheme 3).
Given the ease of preparation of N-aluminoimines and
the possibility of the use of enolizable imines as the starting
materials, we chose to use the N-aluminoimines rather than
N-silylimines for an expanded study of crotylboration and
alkoxyallylboration (see below).
Scheme 4. Allylboration of N-aluminoimines: a) DIBAL-H; Et2O, 1 h,
08C. b) 1) 4a–c, f–i added to I, Et2O, ꢀ100!ꢀ558C; 2) MeOH; 3 h,
ꢀ100!ꢀ558C; 3) NaOH, H2O2; ꢀ788C!RT.
Allylboration of N-aluminoimines: Although we obtained
good results from the reaction of the N-silylaldimines with I
after the addition of water, we
could not extend the reaction
Table 2. Allylboration of N-aluminoimines.
to the preparation of aliphatic
homoallylic amines. We envis-
aged that the use of N-alumi-
noimines could alleviate this
issue as the stability of the N-
aluminoimines is considerably
higher.[13] Itsuno and co-work-
ers have reported the allylbora-
tion of N-aluminoimines.[29]
However, the importance of
methanol or water is not recog-
nized in their account. More-
over, the yields of the amine
products were low and the %
ee were inconsistent with the
reaction temperature. For ex-
Entry
N-Aluminoimine
R
T [8C]
Homoallylic amine
Yield [%][a]
ee [%][b]
1
2
3
4
4
6
7
8
9
7a
7b
7c
7 f
7g
7h
7h
7i
Ph
2-Thp
4-MeO-(C6H4)
4-NO2-(C6H4)
C6F5
nBu
nBu
Chx
Chx
ꢀ100
ꢀ78
2a
2b
2c
2 f
2g
2h
2h
2i
90 (87)[c]
88 (94)[c]
81 (81)[c]
91 (92)[c]
86
74 (72)[c]
ꢀ100
ꢀ100
ꢀ78
89 (74)[c]
85
61
12
65
16
62
79
ꢀ100
ꢀ55
78[d]
60[d]
ꢀ100
ꢀ55
89[d]
7i
2i
68[d]
[a] All yields are of pure isolated products. [b] Enantiomeric excess was determined with HPLC using Chiracel
OD-H column and hexanes/isopropanol as the mobile phase. [c] The values in parentheses are for the amines
obtained from N-silylimines 1a–c. [d] Enantiomeric excess determined by 19F NMR spectroscopy after conver-
sion to Mosher amides; results were confirmed using HPLC analysis.
ample, N-aluminobenzaldimine 7a provided the homoallylic
amine 2a in 59% yield and 33% ee at ꢀ788C and 67%
yield and 69% ee at 258C.[29] This suggested that the reac-
tion could be taking place during the workup. We speculated
that the addition of methanol or water might be equally im-
portant in the allylboration of N-aluminoimines as well.
Partial reduction of benzonitrile (6a) with DIBAL-H in
Et2O at 08C furnished the corresponding N-aluminoimine
7a, which was added to a solution of I in Et2O/pentane at
ꢀ1008C. After several hours, the 11B NMR spectrum of an
aliquot revealed a peak at d 78 corresponding to unchanged
Figure 4. Tentative mechanism for the allylboration of N-aluminoimines.
1
I and H NMR showed the presence of unreacted N-alumi-
noimine (d 9.0), supporting our intuition. Again, addition of
methanol initiated an exothermic reaction and within mi-
nutes the 11B NMR spectrum of an aliquot showed a peak at
d 47, corresponding to an amino dialkylborane species,[30]
while the d 9.0 peak observed in 1H NMR disappeared.
Upon completion of the reaction (within 3 h, based on
11B NMR analysis), oxidative workup furnished the expected
amine 2a in 90% yield and 88% ee, which is considerably
better than the 59% yield and 33% ee reported previously
We believe that the allylboration of imines proceeds anal-
ogous to the allylboration of aldehydes, via a six-membered
chair-like transition state with the stereochemical outcome
determined by the isopinocampheyl auxiliary (Fig-
ure 5).[18a,31] In the cases of the crotylboration, there are
eight possible transition states, four of which will be pre-
dominant, depending on the geometry of the isopinocam-
pheyl group and the butene used.[32] Comparing the optical
rotation values of the obtained amine 2a with those report-
4390
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4387 – 4395

Homoallylic Amines
FULL PAPER
Imine 7a, obtained from benzonitrile (6a), was thus treat-
ed with V or VI in THF at ꢀ788C and after the addition of
1 equivof methanol, followed by the workup provided the
expected amines 3a and 4a, respectively, in good yields and
very high de and ee, similar to the results established from
the reaction of 1a with V and VI (Scheme 5; Table 4, en-
tries 1 and 5). Crotylboration of another aromatic substrate,
Figure 5. Transition states in allylboration of aldimines.
N-alumino-2-thiophenecarbaldimine
(7b)
proceeded
ed in the literature,[33] we assigned its configuration as S. We
assigned the configuration of other homoallylic amines
based on this analogy as well as on the basis of the configu-
ration of several amino acids prepared from these amines
(see below). Configuration of the obtained homoallylic
amine supports the mechanism and the postulated transition
states.
smoothly in THF furnishing the desired products 3b and 4b
(Scheme 5; Table 4, entries 2 and 6). Unfortunately, aliphatic
imines 7h–i did not react so readily at ꢀ788C in THF and
gave only marginal yields of the desired amines along with
several unidentified degradation products. Unlike the reac-
tions of 7h–i with I, increasing the reaction temperature to
ꢀ558C did not improve the yields. Attempted use of catalyt-
ic In(OTf)3 or use of a bulky proton source or water as the
additive did not augment the yield, either. Propitiously, ex-
amination of various solvents revealed that the reaction pro-
ceeded satisfactorily at ꢀ788C when pentane was used as
the solvent for both the preparation of imine 7i and the cro-
tylboration and the homoallylic amine 3i was obtained in
74% yield and >98% diastereoselectivity (Table 3).
After we had obtained the homoallylic amine 2a in high
ee from benzonitrile (6a) via N-aluminoaldimine 7a, we
turned our attention to other classes of nitriles: heteroaro-
matic (2-thiophenecarbonitrile: 6b), electron-donating (4-
methoxybenzonitrile: 6c), electron-withdrawing (4-nitroben-
zonitrile: 6 f), and perfluorinated (pentafluorobenzonitrile:
6g). Additionally, we extended the reaction to aliphatic ni-
triles (valeronitrile: 6h and cyclohexanecarbonitrile: 6i)
(Scheme 4). Thus, treatment of 6a–c and 6 f–i with 1 equiv
of DIBAL-H furnished the desired N-aluminoimines 7a–c
and 7 f–i, which were transferred to a cooled solution of
1.2 equivof I, followed by the addition of 1 equivof metha-
nol (Scheme 4). Allylborations of the aromatic substrates
7a–c and 7 f–g were facile and upon oxidative workup, the
corresponding homoallylic amines 2a–c and 2 f–g were ob-
tained in 61–90% yield and 86–91% ee (as determined by
HPLC analysis; Table 2; entries 1–5). To our dismay, ali-
phatic substrates 7h–i did not provide such good results and
amines 2h–i were obtained only in very poor yields (12–
16%) at ꢀ1008C, while the enantioselectivity remained
good (Table 1, entries 6, 8). Fortunately, reactions run at
higher temperatures afforded better yields and we found
that at ꢀ558C aliphatic homoallylic amines 2h–i were ob-
tained in 62–65% yield and 60–79% ee (Scheme 4; Table 2,
entries 7 and 9). Comparison of the yields, enantioselectivi-
ties, and diastereoselectivities of the homoallylic amines ob-
tained by allylboration of N-silyl- and N-aluminoimines re-
vealed marginal advantage of the N-aluminoimines in terms
of yields, while the N-silylimines furnished amines in slightly
higher ee (Table 2). However, because of the ease of prepa-
ration of the N-aluminoimines from the nitriles, we recom-
mend utilization of this protocol for the preparation of ho-
moallylic amines via allylboration.
Table 3. Effects of temperature, Lewis acids, and solvents on crotylbora-
tion of aliphatic imine 7i with V.
Solvent
Lewis acid
Additive
T [8C]
Yield
[%]
THF
THF
THF
THF
none
none
none
none
MeOH
MeOH
H2O
ꢀ78
ꢀ55
ꢀ78
ꢀ78!
25
21[a]
32[a]
38[a]
27[a]
tBuOH
THF
In(OTf)3
(0.1 equiv)
none
none
none
MeOH
ꢀ78
22[a]
Et2O
MeOH
MeOH
MeOH
H2O
ꢀ78
ꢀ78
ꢀ78
ꢀ78
30[a]
35[a]
38[a]
74
Et2O + pentane[b]
pentane + THF[c]
pentane
none
[a] Pure material was not obtained (50–90% purity by 1H NMR; the im-
purities were not identified). [b] A 1:1 mixture was used for the prepara-
tion of 7i and crotylboration. [c] 7i was prepared in THF and pentane
was used as the solvent for crotylboration.
Hence, crotylboration of 7h–i with V and VI was carried
out in pentane. After the addition of 1 equivof methanol
and completion of the reaction (within 3 h, monitored by
11B NMR spectroscopy), alkaline oxidative workup yielded
b-methyl homoallylic amines 3h–i and 4h–i, respectively, in
good yields (61–78%), excellent diastereoselectivities
Crotylboration and alkoxyallylboration of N-aluminoimines:
Once we had achieved satisfactory crotylboration and alkox-
yallylboration of N-silylimine 1a with the “ate” complexes
V–VII, we initiated a project involving the crotylboration of
N-aluminoimines. Similar to the N-silylimines, the reactions
proceeded only with V or VI. Repeating the blank experi-
ment of mixing 7a with BF3·OEt2 and methanol furnished
the same results as observed in the case of N-silylimine 1a.
(>98%),
and
high
enantioselectivities
(79–89%)
(Scheme 5; Table 4, entries 3–4, 7–8).
Having understood that the alkoxyallylboration of N-sily-
limine 1a required the “ate” complex VII to proceed; we
evaluated this reaction on N-aluminoimines. Accordingly, 7a
was mixed with VII, followed by the addition of 1 equivof
methanol and the reaction was monitored with 11B NMR
spectroscopy. Upon completion (within 3 h, based on
Chem. Eur. J. 2005, 11, 4387 – 4395
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4391

P. V. Ramachandran and T. E. Burghardt
Scheme 6. Alkoxyallylboration of N-aluminoimines. a) 1) VII; THF,
ꢀ788C; 2) 7a–b, h–i; 3) MeOH; 3 h, ꢀ788C; 4) NaOH, H2O2; ꢀ788C!
RT.
Scheme 5. Crotylboration of N-aluminoimines: a) 1) V or VI; THF,
ꢀ788C; 2) 7a–b; b) 1) V or VI; pentane, ꢀ788C; 2) 7h–i; c) MeOH; 3 h,
ꢀ788C; d) NaOH, H2O2; ꢀ788C!RT.
5a with Boc2O, followed by
ozonolysis in CH2Cl2 at ꢀ788C
gave the intermediate alde-
Table 4. Crotylboration of N-aluminoimines.
Entry
Reagent
N-Aluminoimine
Homoallylic amine
hydes, which upon oxidation
R
R’
R’’
Yield [%][a]
de [%][b]
ee [%][c]
with sodium chlorite in 2-
methyl-2-propanol in the pres-
ence of 2-methyl-2-butene and
aqueous NaH2PO4 provided the
desired N-Boc-protected b-
amino acids. The above proce-
dure when applied to 4a and
5a resulted in a loss of diaster-
eoselectivity. Fortunately, the
diastereoselectivity was re-
tained by a direct oxidation of
4a and 5a to acids using NaIO4
1
2
3
4
5
6
7
8
V
V
V
V
VI
VI
VI
VI
7a
7b
7h
7i
7a
7b
7h
7i
Ph
Me
Me
Me
Me
H
H
H
H
H
H
H
H
3a
3b
3h
3i
4a
4b
4h
4i
80 (78)[d]
71
98 (96)[d]
98
>98
89 (92)[d]
89
2-Thp
nBu
Chx
Ph
2-Thp
nBu
Chx
64
74
79[e]
>98
89[e]
H
78 (76)[d]
>98 (98)[d]
>98
90 (90)[d]
88
Me
Me
Me
76
61
78
>98
>98
81[e]
87[e]
[a] All yields are of pure isolated products. [b] Diastereomeric excess was determined by 1H NMR spectrosco-
py. [c] Enantiomeric excess was determined with HPLC using Chiracel OD-H column and hexanes/isopropa-
nol as the mobile phase. [d] The values in parentheses were obtained for the N-silylimine 1a. [e] Enantiomeric
excess determined by H and 19F NMR spectroscopy after conversion to Mosher amides.
1
11B NMR analysis), alkaline ox-
idative workup furnished the b-
Table 5. Alkoxyallylboration of N-aluminoimines.
Entry
N-Aluminoimine
R
Homoallylic amine
alkoxy homoallylic amine 5a
with the results (yield, diaster-
eo- and enantioselectivity) com-
parable with those obtained
from N-silylimine (Scheme 6;
Table 5, entry 1). Fortunately,
unlike in the case of crotylbora-
tion, alkoxyallylboration with
VII proceeded smoothly in
THF at ꢀ788C and after the ad-
dition of methanol the reaction
was complete within 3 h
Yield [%][a]
65 (65)[d]
71
60
62
de [%][b]
ee [%][c]
1
2
3
4
7a
7b
7h
7i
Ph
5a
5b
5h
5i
>98 (98)[d]
95 (95)[d]
87
2-Thp
nBu
Chx
>98
>98
>98
1
92[e]
89[e]
[a] All yields are of pure isolated products. [b] Diastereomeric excess was determined by H NMR spectrosco-
py. [c] Enantiomeric excess was determined with HPLC using Chiracel OD-H column and hexanes/isopropa-
nol as the mobile phase. [d] The values in parentheses were obtained for the alkoxyallylboration of N-silyl-
imine 1a. [e] Enantiomeric excess was determined by 19F NMR spectroscopy after conversion to Mosher
amides.
in the presence of catalytic RuCl3·H2O. Upon deprotection
(11B NMR analysis). After alkaline oxidative workup, the
desired b-alkoxy amines 5b, h, and i were obtained in good
yields (60–71%), excellent de (>98%), and high ee (87–
95%) (Scheme 6; Table 5, entries 2–4).
with ethereal HCl, the desired amino acids were obtained in
good yields (71–82%) as their hydrochloride salts 8a, 8g,
8h, 8a’, and 8a’’ (Scheme 7). Comparison of the optical ro-
Conversion of homoallylic amines to b-amino acids: b-
Amino acid units are present in many natural products with
pharmacological properties and are of utmost importance in
medicinal and biochemical research.[34] For example, the
(2R,3S)-N-benzoyl-3-phenylisoserine moiety is crucial for
the biological activity of Taxol.[34,35] Due to their high pro-
teolytic stability, b-amino acids are outstanding substrates
for peptidomimetics.[36] To demonstrate the application of
our methodology, we converted representative homoallylic
amines to b-amino acids. Protection of 2a, 2g, 2h, 4a, and
Scheme 7. Prepartion of b-amino acids: a) Boc2O; Et2O, 6 h, RT;
b) 1) O3, CH2Cl2, 2 h, ꢀ788C; 2) Me2S; 1 h, ꢀ788C!RT; c) NaClO2,
NaH2PO4; tBuOH, 2-methyl-2-butene, H2O, 0.5 h, RT; d) NaIO4, cat.
RuCl3·H2O; CH3CN/H2O 1:1, 0.5 h, RT; e) HCl; Et2O, 0.5 h, RT.
4392
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4387 – 4395

Homoallylic Amines
FULL PAPER
tation of 8a and 8a’ with the values reported in the litera-
ture indicated that our stereochemistry assignments were
correct.[37] Thus, the tandem allylboration–oxidation of N-
aluminoimines can be used as a general method for the
preparation of optically active b-amino acids from aromatic
or aliphatic nitriles in two steps.
tion of 1 equivof methanol or water is critical for the
“allyl”-boration of both N-silyl- and N-aluminoimines. We
observed that crotylboration and alkoxyallylboration reac-
tions proceed with the “ate” boron complexes rather than
with “allyl”-dialkylborane reagents. We have employed this
methodology for the enantioselective synthesis of represen-
tative b-amino acids, d-amino alcohols, and g-lactams in 2–3
steps from nitriles. We believe that due to the ready availa-
bility or the ease of preparation of nitriles, simplicity of the
reactions, and high reliability in achieving both enantiomers
of the homoallylic amines in excellent diastereo- and enan-
tioselectivities, this methodology will find wide applications
in organic syntheses.
Preparation of d-amino alcohols and g-lactams: To demon-
strate further applications of our protocol, we prepared sev-
eral d-amino alcohols and converted them to g-lactams. g-
Amino acids and g-lactams have found less utility than their
b-analogues in medicinal chemistry. However, they are ex-
cellent peptidomimetics agents and are important parts of
molecules with high biological activity.[38] Hence, we were
attracted to their synthesis. Homoallylic amines 2a, 2h, 4a,
and 5a were N-Boc-protected and hydroborated with 9-
BBN (Chx2BH was used for hydroboration of 5a to simplify
the purification) and oxidized under alkaline conditions to
furnish the corresponding N-Boc-protected d-amino alco-
hols 9a, 9h, 9a’, and 9a’’ in 72–86% yields. These alcohols
were oxidized to N-Boc protected g-amino acids 10a, 10h,
10a’, and 10a’’ by using PDC in DMF in 86–93% yields. Re-
moval of the Boc protection with CF3COOH furnished free
g-amino acids, which under the reaction conditions formed
the corresponding lactams 11a, 11h, 11a’, and 11a’’ in essen-
tially quantitative yields (Scheme 8). Comparison of the op-
tical rotation of 11h with the value reported in the litera-
ture[39] served as an additional proof that the assignment of
the configuration is accurate.
Experimental Section
Representative experimental procedures for the allylboration, crotylbora-
tion, alkoxyallylboration, preparation of b-amino acids, and g-lactams are
given below. Additional procedures and spectral data are available in the
Supporting Information.
N-Trimethylsilylbenzaldimine (1a): n-Butyllithium (2.5m in hexanes;
22 mL, 55 mmol) was slowly added to a solution of 1,1,1,3,3,3-hexame-
thyldisilazide (12 mL, 57 mmol) in Et2O (57 mL) cooled to ꢀ508C and
the reaction was stirred for 0.5 h. Benzaldehyde (5.5 mL, 54 mmol) was
subsequently added and the mixture was stirred for 1 h. Removal of
Et2O under reduced pressure followed by vacuum distillation (b.p. 988C,
1.6 mmHg) furnished 1a (7.8 g, 44 mmol, 81%). 1H NMR (300 MHz,
CDCl3): d = 0.29 (s, 9H), 7.45–7.47 (m, 3H), 7.81–7.82 (m, 2H), 9.00 (s,
1H); 13C NMR (75 MHz, CDCl3): d = ꢀ1.1, 128.5, 128.6, 131.3, 139.4,
168.6.
(1S)-1-Phenylbut-3-en-1-amine (2a) from N-silylimine: Compound 1a
(0.9 g, 5.1 mmol) was added to a stirred solution of (ꢀ)-B-allyldiisopino-
campheylborane (I; 1m in pentane; 6 mL, 6 mmol) diluted with THF
(5 mL) and cooled to ꢀ788C, followed by a slow addition of water
(0.09 mL, 5.0 mmol) in THF (0.5 mL). The mixture was stirred for 1 h at
ꢀ788C and it was oxidised with NaOH (3m in H2O; 2 mL) and (slowly!)
H2O2 (30% in H2O; 1.2 mL) and was left stirring under positive N2 pres-
sure while it slowly warmed to RT. The product was then extracted with
Et2O (3”50 mL), treated with HCl (20% in H2O; 5 mL), and stirred for
0.2 h. To the mixture was added water (50 mL) to extract the product.
After removal of the organic layer, the aqueous solution of amine hydro-
chloride was neutralised with NaOH until pH ~8. The resulting amine
was extracted with Et2O (3”50 mL), the solvent was removed under re-
duced pressure, and the material was purified on silica gel (hexanes/ethyl
acetate/triethylamine 84.5:15:0.5) to afford 2a (0.66 g, 4.5 mmol, 90%) in
1
92% ee (HPLC: Chiracel OD-H column, hexanes/isopropanol). H NMR
(300 MHz, CDCl3): d = 1.69 (brs, 2H), 2.32–2.50 (m, 2H), 4.00 (d, J=
8.0 Hz, 1H), 5.07–5.15 (m, 2H), 5.69–5.82 (m, 1H), 7.22–7.37 (m, 5H);
13C NMR (75 MHz, CDCl3): d = 44.2, 55.4, 117.7, 126.4, 127.0, 128.5,
135.5, 145.8; MS (EI): m/z: 128, 106 [Ph-CH-NH2]+, 79; (CI): m/z: 148
[M+H]+, 131 [MꢀNH3]+; HRMS: m/z: calcd for: 148.1126, found:
148.1129; [a] = +43 (CHCl3, c=1.9), lit.:[33] +42, CHCl3, c=0.5).
Scheme 8. Preparation of d-amino alcohols and g-lactams: a) Boc2O;
Et2O, 4 h, RT; b) 1) 9-BBN or Chx2BH; THF, 24 h, RT; 2) NaOAc or
NaOH, H2O2; 5 h, RT. c) PDC; DMF, 12 h, RT; d) CF3COOH; CH2Cl2,
0.5 h, RT. Chx = Cyclohexyl.
(1S)-1-Phenylbut-3-en-1-amine (2a) from benzonitrile: DIBAL-H
Conclusion
(0.89 mL, 5.0 mmol) was added to
a solution of benzonitrile (6a;
0.52 mL, 5.05 mmol) in Et2O (5 mL) cooled to 08C and the mixture was
stirred for 1 h to obtain 7a. 1H NMR (300 MHz, CDCl3): d = 0.14–0.19
(m, 3H), 0.76–1.07 (m, 12H), 1.79 (qn, J=6.6 Hz, 3H), 7.48–7.80 (m,
5H), 9.00 (s, 1H); 13C NMR (75 MHz, CDCl3): d = 22.7, 22.8, 26.3, 26.5,
28.2, 28.3, 28.5, 28.7, 129.4, 129.5, 132.4, 132.8, 133.0, 137.1, 174.5, 175.0).
In conclusion, we have demonstrated the utility of “allyl”-
boration of N-silylimines and N-aluminoimines for the prep-
aration of homoallylic amines in high enantiomeric excess.
For the first time, we have reported the reagent-controlled
crotylboration and alkoxyallylboration of N-masked imines
for the synthesis of b-substituted homoallylic amines in ex-
cellent de and high ee. We have demonstrated that the addi-
Compound 7a was transferred via cannula to a solution of I (1m in pen-
tane; 6 mL, 6 mmol) diluted with Et2O (7 mL) and cooled to ꢀ1008C,
followed by a slow addition of methanol (0.20 mL, 5.0 mmol). The mix-
ture was stirred for 3 h, while it was allowed to slowly warm from ꢀ100
Chem. Eur. J. 2005, 11, 4387 – 4395
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4393

P. V. Ramachandran and T. E. Burghardt
to ꢀ788C and it was oxidised with NaOH (3m in H2O; 2 mL) and
(slowly!) H2O2 (30% in H2O; 1.2 mL) and was left stirring under positive
N2 pressure while it slowly warmed to RT. The product was then extract-
ed with Et2O (3”50 mL), treated with HCl (20% in H2O; 5 mL), and
stirred for 0.2 h. To the mixture was added water (50 mL) to extract the
product. After removal of the organic layer, the aqueous solution of
amine hydrochloride was neutralised with NaOH until pH ~8. The re-
sulting amine was extracted with Et2O (3”50 mL), the solvent was re-
moved under reduced pressure, and the material was purified on silica
gel (hexanes/ethyl acetate/triethylamine 84.5:15:0.5) to afford 2a (0.66 g,
4.5 mmol, 90%) in 88% ee as analysed by the HPLC, having identical
0.5 h at ꢀ788C. Then, a solution of (ꢀ)-B-methoxydiisopinocampheylbor-
ane (2.37 g, 7.5 mmol) in THF (5 mL) was added and the mixture was
stirred for 1 h. To thus generated VII was added via cannula a solution of
7a [prepared as follows: To 6a (0.52 mL, 5.05 mmol) diluted with THF
(5 mL) and cooled to 08C was added DIBAL-H (0.89 mL, 5.0 mmol) and
the mixture was stirred for 1 h], followed by methanol (0.20 mL,
5.0 mmol). The reaction was stirred for 3 h at ꢀ788C and was oxidised
with NaOH (3m in H2O; 2 mL) and (slowly!) H2O2 (30% in H2O;
1.2 mL). The material was left stirring under positive N2 pressure while it
slowly warmed to RT. The product was extracted with Et2O (3”50 mL)
and the volatiles were removed under reduced pressure. The obtained
material was purified on silica gel (hexanes/ethyl acetate/triethylamine
94.5:5:0.5 to 69.5:30:0.5) to furnish 5a (0.81 g, 3.2 mmol, 65%) in >98%
de (1H NMR analysis) and 95% ee (HPLC: Chiracel OD-H, hexanes/iso-
spectral data to the reported above; [a]=+39 (CHCl3, c=0.10), lit.:[33]
42 (CHCl3, c=0.5).
+
(1S,2S)-2-Methyl-1-phenylbut-3-en-1-amine (3a): trans-Butene (1 mL,
10 mmol) and n-butyllithium (2.5m in hexanes; 2.4 mL, 6.0 mmol) were
added to potassium tert-butoxide (1m in THF; 6 mL, 6 mmol) diluted
with THF (6 mL) and cooled to ꢀ788C. The mixture was stirred for 0.1 h
at ꢀ788C, followed by 0.3 h at ꢀ558C, and cooled again to ꢀ788C, when
a solution of (ꢀ)-B-methoxydiisopinocampheylborane (2.28 g, 7.2 mmol)
in THF (5 mL) was added and the reaction was stirred for 1 h at ꢀ788C.
To thus generated V was added via cannula a solution of 7a [prepared as
follows: To 6a (0.52 mL, 5.05 mmol) diluted with THF (5 mL) and
cooled to 08C was added DIBAL-H (0.89 mL, 5.0 mmol) and the mixture
was stirred for 1 h], followed by a slow addition of methanol (0.20 mL,
5.0 mmol) and the mixture was stirred for 3 h at ꢀ788C, when it was oxi-
dized with NaOH (3m in H2O; 2 mL) and (slowly!) H2O2 (30% in H2O;
1.2 mL) and was left stirring under positive N2 pressure while it slowly
warmed to RT. The product was extracted with Et2O (3”50 mL) after
the acid-base manipulation, the solvent was removed under reduced pres-
sure, and the crude material was purified on silica gel (hexanes/ethyl ace-
tate/triethylamine 84.5:15:0.5) to afford 3a (0.59 g, 3.7 mmol, 74%).
1
propanol). H NMR (300 MHz, CDCl3): d = 1.73 (brs, 2H), 3.34 (s, 3H),
3.37–3.49 (m, 4H), 3.96 (d, J=5.8 Hz, 1H), 4.18 (t, J=6.5 Hz, 1H), 4.67
(dd, J=6.9 Hz, 38.7 Hz, 2H), 5.11–5.11 (m, 2H), 5.58–5.69 (m, 1H),
7.21–7.34 (m, 5H); 13C NMR (75 MHz, CDCl3): d
= 59.0, 59.8, 67.0,
71.7, 81.7, 93.0, 118.7, 127.2, 127.5, 128.2, 135.5, 142.6; MS (EI): m/z: 176
[MꢀOCH2CH2OCH3]+, 106, 79, 59; (CI): m/z: 252 [M+H]+, 176
[M+HꢀCH3OCH2CH2OH]+, 106, 79; HRMS: m/z: calcd for 252.1600,
found: 252.1604; [a]=+103 (CHCl3, c=4.22).
(2R,3S)-3-Amino-2-methyl-3-phenylpropanoic acid hydrochloride (8a’):
Di-tert-butyl dicarbonate (0.3 g, 1.4 mmol) was added to 4a (0.2 g,
1.2 mmol) dissolved in Et2O (12 mL) and the reaction was stirred for 6 h
at RT, after which time the solvent was removed under reduced pressure.
To the crude material dissolved in CH3CN (40 mL) was added
RuCl3·H2O (0.02 g, 0.1 mmol) and the mixture was cooled to 08C. After
addition of NaIO4 (0.8 g, 3.7 mmol) dissolved in water (40 mL), the mix-
ture was stirred for 0.5 h, followed by extraction with EtOAc (3”30 mL)
and filtration through silica gel (Et2O). After evaporation of the solvents,
the residue was diluted with Et2O (5 mL) and treated with HCl (1m in
Et2O; 2 mL, 2 mmol) for 0.5 h. The obtained solid was filtered and dried
to afford 8a’ (0.2 g, 0.9 mmol, 77%). 1H NMR (200 MHz, D2O): d =
0.91 (d, J=7.2 Hz, 3H), 2.97–3.09 (m, 1H), 4.36 (d, J=9.4 Hz, 1H), 7.29–
1H NMR (300 MHz, CDCl3): d
= 0.83 (d, J=6.7 Hz, 3H), 1.53 (brs,
2H), 2.37 (q, J=7.4 Hz, 1H), 3.65 (d, J=8.4 Hz, 1H), 5.10–5.24 (m, 2H),
5.69–5.81 (m, 1H), 7.26–7.33 (m, 5H); 13C NMR (75 MHz, CDCl3): d =
17.7, 46.4, 60.7, 115.9, 127.1, 127.3, 128.3, 141.8, 144.7; MS (EI): m/z: 160
[MꢀH]+, 106, 79; (CI): m/z: 162, 145, 106; [a]=+76 (CHCl3, c=0.92),
lit:[37b] +1.5 (MeOH, c=1.0).
7.31 (m, 5H); 13C NMR (50 MHz, D2O): d
= 15.4, 43.9, 57.8, 127.4,
127.6, 129.6, 129.8, 134.5, 177.3; [a]=+19 (D2O, c=1.19).
(5S)-5-Phenylpyrrolidin-2-one (11a): Di-tert-butyl dicarbonate (0.6 g,
3.1 mmol) was added to 2a (0.43 g, 2.9 mmol) dissolved in Et2O (30 mL)
and the reaction was stirred for 6 h at RT, after which time the solvent
was removed under reduced pressure. The crude material was dissolved
in THF (7 mL) and treated with 9-BBN (0.5m in THF; 13 mL, 6.5 mmol)
for 24 h at RT, followed by oxidation with sodium acetate (20% in H2O,
20 mL) and H2O2 (30% in H2O; 6 mL) for 5 h at RT. The product was
extracted with Et2O (3”30 mL), and after evaporation of the solvents pu-
rified on silica gel (hexanes/ethyl acetate 2:1) to furnish tert-butyl (1S)-4-
hydroxy-1-phenylbutylcarbamate (9a) (0.66 g, 2.5 mmol, 86%). 1H NMR
(300 MHz, CDCl3): d = 1.45 (s, 9H), 1.53–1.88 (m, 4H), 2.55 (brs, 1H),
3.67 (t, J=5.9 Hz, 2H), 4.67 (brs, 1H), 5.12 (brs, 1H), 7.29–7.39 (m,
5H); 13C NMR (75 MHz, CDCl3): d = 28.3, 29.1, 33.2, 54.6, 62.1, 79.5,
126.3, 127.2, 128.5, 142.7, 155.5.
(1R,2S)-1-Butyl-2-methylbut-3-enylamine (3h): trans-Butene (1 mL,
10 mmol) and n-butyllithium (2.5m in hexanes; 2.4 mL, 6.0 mmol) were
added to potassium tert-butoxide (1m in THF; 6 mL, 6 mmol) diluted
with pentane (6 mL) and cooled to ꢀ788C. The mixture was stirred for
0.1 h at ꢀ788C, followed by 0.3 h at ꢀ558C, and cooled again to ꢀ788C,
when
a
solution of (ꢀ)-B-methoxydiisopinocampheylborane (2.28 g,
7.2 mmol) in pentane (5 mL) was added and the reaction was stirred for
1 h at ꢀ788C. To thus generated V was added via cannula a solution of
7h [prepared as follows: To a solution of valeronitrile (6h) (0.55 mL,
5.2 mmol) in pentane (10 mL) cooled to 08C was added DIBAL-H
(0.90 mL, 5.0 mmol) and the mixture was stirred for 1 h], followed by
methanol (0.20 mL, 5.0 mmol) and the mixture was stirred for 3 h at
ꢀ788C when it was oxidised with NaOH (3m in H2O; 2 mL) and
(slowly!) H2O2 (30% in H2O; 1.2 mL) and was left stirring under positive
N2 pressure while it slowly warmed to RT. The product was extracted
with Et2O (3”50 mL) after the acid/base manipulations, the solvent was
removed under reduced pressure, and the crude material was purified on
silica gel (hexanes/ethyl acetate/triethylamine 84.5:15:0.5) to afford 3h
(0.4 g, 2.8 mmol, 64%) in >98% de (1H NMR analysis) and 79% ee
The N-Boc protected alcohol 9a (0.22 g, 0.8 mmol) in DMF (10 mL) was
added to a stirring solution of pyridinium dichromate (1.13 g, 3.0 mmol)
in DMF (20 mL) and the mixture was stirred for 18 h at RT. The reaction
was quenched with H2O (5 mL), the product was extracted with Et2O
(3”50 mL), the combined ether layers were washed with H2O (3”
50 mL), the solvent was removed and the obtained material was purified
on silica gel (flash; hexanes/ethyl acetate 2:1) to afford (4S)-4-[(tert-bu-
toxycarbonyl)amino]-4-phenylbutanoic acid (10a) (0.192 g, 0.7 mmol,
1
(analysis using H and 19F NMR spectroscopy after conversion to Mosher
amide and confirmation using HPLC with Chiracel OD-H column with
hexanes/isopropanol). 1H NMR (200 MHz, CDCl3): d
= 0.95 (t, J=
1
4.5 Hz, 3H), 1.06 (d, J=4.6 Hz, 3H), 1.27–1.40 (m, 8H), 2.16 (q, J=
4.5 Hz, 1H), 2.57–2.60 (m, 1H), 5.06–5.09 (m, 2H), 5.72–5.84 (m, 1H);
13C NMR (50 MHz, CDCl3): d = 14.4, 17.0, 23.2, 28.9, 34.9, 44.3, 55.6,
115.4, 141.4. MS (EI): m/z: 142 [M+H]+, 86 [MꢀC4H7]+; (CI): m/z: 142
[M+H]+, 86; [a]=+15 (CDCl3, c=1.28).
86%). H NMR (200 MHz, CDCl3): d = 1.22 (s, 9H), 1.78–2.01 (m, 1H),
2.35–2.71 (m, 3H), 5.07–5.13 (m, 1H), 7.15–7.34 (m, 5H); 13C NMR
(50 MHz, CDCl3): d = 28.0, 28.3, 31.9, 61.9, 82.9, 124.8, 127.3, 128.4,
142.1, 149.0, 174.1.
The N-Boc protected g-amino acid 10a (0.13 g, 0.5 mmol) dissolved in
CH2Cl2 (3 mL) was treated with CF3COOH (0.1 mL) for 0.5 h at RT.
After evaporation, the obtained material was purified on silica gel (flash;
hexanes/ethyl acetate 1:1) to give 11a (0.074 g, 0.5 mmol, 98%).
1H NMR (300 MHz, CDCl3): d = 1.97–2.11 (m, 1H), 2.41–2.70 (m, 3H),
(1R,2R)-2-[(2-Methoxyethoxy)methoxy]-1-phenylbut-3-en-1-amine (5a):
sec-Butyllithium (1.4m in cyclohexane; 4.4 mL, 6.1 mmol) was added to
3-[(2-methoxyethoxy)-methoxy]prop-1-ene (0.91 g, 6.2 mmol) diluted
with THF (6 mL) and cooled to ꢀ788C and the mixture was stirred for
4394
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4387 – 4395

Homoallylic Amines
FULL PAPER
4.83 (t, J=7.1 Hz, 1H), 6.70 (brs, 1H), 7.34–7.46 (m, 5H); 13C NMR
(75 MHz, CDCl3): d = 30.3, 31.3, 58.1, 125.6, 127.8, 128.8, 142.5, 178.7;
MS (EI): m/z: 161 [M]+, 117, 104, 77; (CI): m/z: 162 [M+H]+; [a]=+25
(CDCl3, c=3.7).
[18] a) R. W. Hoffmann, H. J. Zeiss, J. Org. Chem. 1981, 46, 1309;
b) H. C. Brown, P. K. Jadhav, J. Am. Chem. Soc. 1983, 105, 2092;
c) H. C. Brown, K. S. Bhat, J. Am. Chem. Soc. 1986, 108, 293;
d) H. C. Brown, P. K. Jadhav, K. S. Bhat, J. Am. Chem. Soc. 1988,
110, 1535.
[19] a) P. V. Ramachandran, Aldrichimica Acta 2002, 35, 23; b) P. V.
Ramachandran, M. V. R. Reddy, H. C. Brown, Pure Appl. Chem.
2003, 75, 1263.
[20] a) S. Itsuno, K. Watanabe, K. Ito, A. A. El-Shehawy, A. A. Sarhan,
Angew. Chem. 1997, 109, 105; Angew. Chem. Int. Ed. Engl. 1997, 36,
109; b) S. Itsuno, K. Watanabe, T. Matsumoto, S. Kuroda, A. Yokoi,
A. A. El-Shehawy, J. Chem. Soc. Perkin Trans. 1 1999, 2011.
[21] H. C. Brown, U. S. Racherla, P. J. Pellechia, J. Org. Chem. 1990, 55,
1868.
Acknowledgements
Financial assistance from the Herbert C. Brown Center for Borane Re-
search[40] and Aldrich Chemical Company is gratefully acknowledged. Dr.
G.-M. Chenꢁs initial work on the allylboration of N-silylimines[22] is ac-
knowledged.
[22] G.-M. Chen, P. V. Ramachandran, H. C. Brown, Angew. Chem. 1999,
111, 828; Angew. Chem. Int. Ed. 1999, 38, 825.
[1] a) C. O. Puentes, V. Kouznetsov, J. Heterocycl. Chem. 2002, 39, 595;
b) J. Cossy, C. Willis, V. Bellosta, S. BouzBouz, J. Org. Chem. 2002,
67, 1982; c) M. Brands, R. Endermann, R. Gahlmann, J. Krüger, S.
Raddatz, Bioorg. Med. Chem. Lett. 2003, 13, 241; d) K. J. Hale,
M. M. Domostoj, D. A. Tocher, E. Irving, F. Scheinmann, Org. Lett.
2003, 5, 2927.
[2] Among the most recent reports, see: a) C. Ogawa, M. Suguiura, S.
Kobayashi, Angew. Chem. 2004, 116, 6653; Angew. Chem. Int. Ed.
2004, 43, 6491; b) M. Sugiura, K. Hirano, S. Kobayashi, J. Am.
Chem. Soc. 2004, 126, 7182; c) P. C. Andrews, A. C. Peatt, C. L.
Raston, Green Chem. 2004, 6, 119; d) B. M. Choudary, K. Jyothi, S.
Madhi, M. L. Kantam, Synlett 2004, 231.
[3] For reviews, see: a) Comprehensive Organic Synthesis, Vol. 2 (Eds.:
B. Trost, I. Fleming), Pergamon Press, Oxford, 1994; b) J. P. Adams,
Contemp. Org. Synth. 1997, 4, 517; c) D. Enders, U. Reinhold, Tetra-
hedron: Asymmetry 1997, 8, 1895; d) R. Bloch, Chem. Rev. 1998, 98,
1407; e) S. Kobayashi, H. Ishitani, Chem. Rev. 1999, 99, 1069.
[4] S. M. Weinreb, Top. Curr. Chem. 1997, 190, 131.
[5] a) F. A. Davis, B.-C. Chen, Chem. Soc. Rev. 1998, 27, 13; b) J. A.
Ellman, T. D. Owens, T. P. Tang, Acc. Chem. Res. 2002, 35, 984.
[6] M. Panunzio, P. Zarantonello, Org. Process Res. Dev. 1998, 2, 49.
[7] a) R. W. Hoffmann, G. Eichler, A. Endesfelder, Liebigs Ann. Chem.
1983, 2000; b) R. W. Hoffmann, A. Endesfelder, Liebigs Ann. Chem.
1987, 215.
[23] G.-M. Chen, H. C. Brown, J. Am. Chem. Soc. 2000, 122, 4217.
[24] Among the most recent reports, see: a) P. V. Ramachandran, B.
Prabhudas, D. Pratihar, J. S. Chandra, M. V. R. Reddy, Tetrahedron
Lett. 2003, 44, 3745; b) K. C. Nicolaou, K. C. Fylaktakidou, H. Mon-
enschein, Y. W. Li, B. Weyershausen, H. J. Mitchell, H. X. Wei, P.
Guntupalli, D. Hepworth, K. Sugita, J. Am. Chem. Soc. 2003, 125,
15433; c) P. V. Ramachandran, B. Prabhudas, J. S. Chandra, M. V. R.
Reddy, J. Org. Chem. 2004, 69, 6294; d) X. Zeng, F. Zeng, E.-I. Ne-
gishi, Org. Lett. 2004, 6, 3245; e) A. V. Statsuk, D. Liu, S. A.
Kozmin, J. Am. Chem. Soc. 2004, 126, 9546.
[25] K. Fujita, M. Schlosser, Helv. Chim. Acta 1982, 65, 1258.
[26] H. C. Brown, J. A. Sinclair, J. Organomet. Chem. 1977, 131, 163.
[27] P. V. Ramachandran, K. J. Padiya, V. Rauniyar, M. V. R. Reddy,
H. C. Brown, Tetrahedron Lett. 2004, 45, 1015.
[28] Among the most recent reports, see: a) P. V. Ramachandran, J. S.
Chandra, M. V. R. Reddy, J. Org. Chem. 2002, 67, 7547; b) X. Liao,
Y. Wu, J. K. De Brabander, Angew. Chem. 2003, 115, 1686; Angew.
Chem. Int. Ed. 2003, 42, 1648; ; c) K. D. Carter, J. S. Panek, Org.
Lett. 2004, 6, 55; d) B. Liu, W.-S. Zhou, Org. Lett. 2004, 6, 71.
[29] K. Watanabe, S. Kuroda, A. Yokoi, K. Ito, S. Itsuno, J. Organomet.
Chem. 1999, 581, 103.
[30] H. Nçth, B. Wrackmeyer, Nuclear Magnetic Resonance Spectroscopy
of Boron Compounds, Springer, Berlin 1978.
[31] H. C. Brown, P. K. Jadhav, J. Org. Chem. 1984, 49, 4089.
[32] H. C. Brown, K. S. Bhat, J. Am. Chem. Soc. 1986, 108, 5919.
[33] T. Basile, A. Bocoum, D. Savoia, A. Umani-Ronchi, J. Org. Chem.
1994, 59, 7766.
[8] G. Cainelli, M. Panunzio, P. Andreoli, G. Martelli, G. Spunta, D.
Giacomini, E. Bandini, Pure Appl. Chem. 1990, 62, 605.
[9] C. Krüger, E. G. Rochow, U. Wannagat, Chem. Ber. 1964, 97, 2132.
[10] D. J. Hart, K.-I. Kanai, D. G. Thomas, T.-K. Yang, J. Org. Chem.
1983, 48, 289.
[34] Enantioselective Synthesis of b-Amino Acids (Ed.: E. Juaristi),
Wiley-VCH 1997 and references therein.
[11] J.-C. Guillemin, L. Ammi, J.-M. Denis, Tetrahedron Lett. 1988, 29,
1287.
[12] G. Cainelli, D. Giacomini, M. Panunzio, G. Martelli, G. Spunta, Tet-
rahedron Lett. 1987, 28, 5369.
[13] P. Andreoli, L. Billi, G. Cainelli, M. Panunzio, G. Martelli, G.
Spunta, J. Org. Chem. 1990, 55, 4199.
[35] I. Ojima, S. N. Lin, T. Wang, Curr. Med. Chem. 1999, 6, 927.
[36] D. L. Steer, R. A. Lew, P. Perlmutter, A. I. Smith, M.-I. Aguilar,
Curr. Med. Chem. 2002, 9, 811.
[37] a) S. Laschat, H. Kunz, J. Org. Chem. 1991, 56, 5883; b) S. G.
Davies, O. Ichihara, A. S. Walters, J. Chem. Soc. Perkin Trans. 1
1994, 1143.
[14] P. V. Ramachandran, M. T. Rudd, T. E. Burghardt, M. V. R. Reddy,
J. Org. Chem. 2003, 68, 9310.
[15] Y. Yamamoto, W. Ito, K. Maruyama, J. Chem. Soc. Chem. Commun.
1985, 1131.
[16] a) Y. Yamamoto, T. Komatsu, K. Maruyama, J. Am. Chem. Soc.
1984, 106, 5031; b) Y. Yamamoto, W. Ito, K. Maruyama, J. Chem.
Soc. Chem. Commun. 1985, 1131; c) Y. Yamamoto, S. Nishi, K. Mar-
uyama, T. Komasu, W. Ito, J. Am. Chem. Soc. 1986, 108, 7778; d) Y.
Yamamoto, N. Asao, Chem. Rev. 1993, 93, 2207.
[38] a) J. J. W. Duan, L. H. Chen, Z. R. Wasserman, Z. H. Lu, R. Q. Liu,
M. B. Covington, M. X. Qian, K. D. Hardman, R. L. Magolda, R. C.
Newton, D. D. Christ, R. R. Wexler, C. P. Decicco, J. Med. Chem.
2002, 45, 4954; b) K. Dolbeare, G. F. Pontoriero, K. S Gupta, R. K.
Mishra, R. L. Johnson, J. Med. Chem. 2003, 46, 727.
[39] S. Saliou, A. Fleurant, J. P. Cꢂlꢂrier, G. Lhommet, Tetrahedron Lett.
1991, 32, 3365.
[40] Publication number 36 from the Herbert C. Brown Center for
Borane Research.
[17] S. Itsuno, C. Hachisuka, K. Kitano, K. Ito, Tetrahedron Lett. 1992,
33, 627.
Received: December 16, 2004
Published online: May 6, 2005
Chem. Eur. J. 2005, 11, 4387 – 4395
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4395